Download pombe fbp1 gene occurs by a cAMP signaling pathway. Glucose

Downloaded from www.genesdev.org on September 10, 2007
Glucose repression of transcription of the Schizosaccharomyces
pombe fbp1 gene occurs by a cAMP signaling pathway.
C S Hoffman and F Winston
Genes & Dev. 1991 5: 561-571
Access the most recent version at doi:10.1101/gad.5.4.561
References
This article cites 56 articles, 23 of which can be accessed free at:
http://www.genesdev.org#References
Article cited in:
http://www.genesdev.org/cgi/content/abstract/5/4/561#otherarticles
Email alerting
service
Receive free email alerts when new articles cite this article - sign up in the box at
the top right corner of the article or click here
Notes
To subscribe to Genes and Development go to:
http://www.genesdev.org/subscriptions/
© 1991 Cold Spring Harbor Laboratory Press
Downloaded from www.genesdev.org on September 10, 2007
Glucose repression of transcription of the
Schizosaccharomyces pombe fbpl gene
occurs by a cAMP signaling pathway
Charles S. Hoffman 1 and Fred W i n s t o n
Department of Genetics, Harvard Medical School, Boston, Massachusetts 02115 USA
Transcription of the fbpl gene, encoding fructose-l,6-bisphosphatase, of Schizosaccharomyces pombe is
subject to glucose repression. Previous work has demonstrated that several genes {g/t genes} are required for
this repression. In this report we demonstrate that one of these genes, git2, is the same as the cyrl gene,
which encodes adenylate cyclase, and that loss-of-function mutations in git2 cause constitutive fbpl
transcription. Addition of cAMP to the growth medium suppresses the transcriptional defect in git2 mutants
as well as in strains that carry mutations in any of six additional g/t genes. Similarly, exogenous cAMP
represses [bpl transcription in wild-type cells grown on a derepressing carbon source. Different levels of
adenylate cyclase activity in different git2 mutants, coupled with the result that some git2 mutants display
intragenic complementation, strongly suggest that adenylate cyclase acts as a multimer and that different git2
mutations alter distinct activities of adenylate cyclase, including catalytic activity and response to glucose.
Additional experiments demonstrate that this cAMP signaling pathway is independent of the S. pombe rasl
gene and works by activation of cAMP-dependent protein kinase.
[Key Words: Schizosaccharomyces pombe; fructose-bisphosphatase; glucose repression; cAMP; adenylate
cyclase; transcription]
Received January 4, 1991; accepted January 15, 1991.
Cells regulate transcription in response to signals from
their environment. For example, unicellular organisms
transcribe specific genes to metabolize the carbon
sources available. If glucose is also present, however,
many genes required for the utilization of other carbon
sources are no longer expressed. This phenomenon has
been termed glucose repression or catabolite repression
{Magasanik 1961).
Glucose repression appears to work through different
mechanisms in the bacterium Escherichia coli and the
yeast Saccharomyces cerevisiae. In E. coil, glucose
causes a reduction in the level of cAMP that binds to and
is required for activity of the catabolite gene activator
protein [for review, see Botsford 1981}. In S. cerevisiae,
however, cAMP is not required for expression of genes
subject to glucose repression {Matsumoto et al. 1982b,
1983) and cAMP levels are not reduced under conditions
that derepress expression of such genes {Eraso and
Gancedo 1984}. In contrast, S. cerevisiae ADH2 transcription appears to be repressed by cAMP through the
phosphorylation and inactivation of the transcriptional
activator ADR1 by cAMP-dependent protein kinase
(cAPK; Cherry, et al. 1989; Taylor and Young 1990). In S.
cerevisiae, glucose repression involves a variety of mechanisms, as different mutations have been isolated that
1Present address: Department of Biology, Boston College, Chestnut Hill,
Massachusetts 02167 USA.
allow constitutive expression of distinct subsets of glucose repressible genes (for review, see Gancedo and
Gancedo 1986; Johnston 1987).
In the yeast Schizosaccharomyces pombe, the [bpl
gene, encoding the gluconeogenic enzyme fructose-l,6bisphosphatase, is subject to glucose repression (Vassarotti and Friesen 1985; Hoffman and Winston 1989,
1990). We have observed that the level of [bpl transcription can vary over a range of >200-fold with respect to
the carbon source. Using a pair of translational fusions to
fbpl, we have isolated mutants in which transcription of
an [bpl-ura4 fusion is constitutive. These mutations
identified 10 genes, designated git [glucose-_insensitive
transcription} [Hoffman and Winston 1990]. Mutations
in 8 of these 10 genes cause a significant increase in [bpl
transcription under normally repressing conditions (see
Fig. 2, below; Hoffman and Winston 1990}. We have begun a molecular analysis of glucose repression by cloning
and analyzing one of these git genes.
In this report we demonstrate that the git2 gene, mutations in which allow high-level constitutive fbpl transcription, is identical to the cyrl gene, which encodes
adenylate cyclase (Yamawaki-Kataoka et al. 1989; Young
et al. 1989). This enzyme functions to convert ATP to
cAMP in response to external stimuli and is involved in
many signaling pathways in eukaryotic cells {Gilman
1984; Levitzki 1988). In these signaling pathways the
role of cAMP is to activate cAPK, which is composed of
GENES & DEVELOPMENT 5:561-571 © 1991 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/91 $3.00
561
Downloaded from www.genesdev.org on September 10, 2007
Hoffman and Winston
a pair of regulatory subunits associated with a pair of
catalytic subunits, cAMP binds the regulatory subunits
of cAPK, causing them to dissociate from and thereby
activate the catalytic subunits of cAPK, which are now
able to phosphorylate target proteins (Beebe and Corbin
1986; Edelman et al. 1987). Some known target proteins
are enzymes that are inactivated by phosphorylation,
whereas others are transcriptional activators (Beebe and
Corbin 1986; Roesler et al. 1988).
In S. cerevisiae, many genes have been identified that
are involved in the adenylate cyclase pathway. Adenylate cyclase is encoded by the CYR1 gene (Casperson et al.
1985), whereas the upstream activators are encoded by
RAS1 and RAS2 (Broek et al. 1985; Toda et al. 1985).
CDC25 encodes a positive regulator of RAS activity required for adenylate cyclase activation (Broek et al. 1987;
Robinson et al. 1987), whereas IRA1 (Tanaka et al. 1989)
and IRA2 (Tanaka et al. 1990) encode negative regulators
of RAS activity. The CYR1 gene product is associated
with a 70-kD protein encoded by the CAP/SRV2 gene
(Fedor-Chaiken et al. 1990; Field et al. 1990). The 70-kD
protein is also required for RAS activation of adenylate
cyclase. The regulatory subunit of cAPK is encoded by
BCY1, mutations in which bypass the requirement for
active adenylate cyclase (Matsumoto et al. 1982a). The
catalytic subunits of cAPK are encoded by each of the
TPK1, TPK2, and TPK3 genes (Toda et al. 1987b).
In S. pombe, less is known about the adenylate cyclase
pathway. The genes that encode adenylate cyclase (cyrl;
Yamawaki-Kataoka et al. 1989; Young et al. 1989) and
the regulatory subunit of cAPK (cgsl; M. McLeod, pers.
comm.) have been identified. An S. pombe RAS gene
homolog, rasl, has been identified (Fukui and Kaziro
1985; Nadin-Davis et al. 1986a); however, in contrast to
S. cerevisiae, rasl does not appear to stimulate adenylate
cyclase activity in S. pombe (Fukui et al. 1986; NadinDavis et al. 1986b; for review, see Nadin-Davis et al. 1989).
Additional experiments in this report demonstrate
that git2 (cyrl), other git genes, and cgsl are required for
normal glucose-mediated regulation of fbpl transcription. As expected, rasl does not play a role in fbpl regulation. In addition, intragenic complementation between various git2 mutant alleles and differential effects
on adenylate cyclase activity (as assayed in vitro) and
cAMP levels by these mutations indicate that adenylate
cyclase is a multimer and that git2 mutations cause constitutive fbpI transcription by altering distinct functions
of the protein. These results lead to the conclusion that
fbpl transcription is regulated by a cAMP signaling pathway, independent of rasl, and that the role of the cAMP
signal in S. pombe glucose repression is to activate
cAPK.
Results
Cloning of the git2 gene as a high copy suppressor of a
girl mutation
Previous work had identified a large number of git genes
required for glucose repression of fbpl transcription
(Hoffman and Winston 1990). To initiate a molecular
562
GENES & DEVELOPMENT
analysis of the git genes, we set out to clone the gitl
gene. Mutations in gitl allow constitutive transcription
of the wild-type fbpl gene, as well as constitutive expression of both fbpl-ura4 and fbpl-lacZ translational
fusions, under normally repressing conditions (Hoffman
and Winston 1990). Because cells that express the
fbpl-ura4 fusion are sensitive to the pyrimidine analog
5-fluoro-orotic acid (SFOA; Boeke et al. 1984; Hoffman
and Winston 1990), g i t l - mutants that contain the
fbpl-ura4 fusion are 5FOA sensitive (SFOAS; G i t - phenotype) when grown under repressing conditions (8%
glucose as the carbon source); gitl + strains are 5FOA
resistant (5FOAR; Git + phenotype). We therefore
screened an S. pombe genomic library in a high copy
number vector for plasmids that could transform a gitl fbpl-ura4 strain (Git-) to Git + (see Materials and methods). After screening 17,000 transformants, we obtained
one plasmid (pCHY26) that conferred a strong Git +
(5FOA R) phenotype.
Three pieces of evidence demonstrated that pCHY26
encodes git2 rather than gitl. First, when pCHY26 was
integrated into a gitI - strain, the integrant unexpectedly
remained G i t - , demonstrating that the cloned DNA,
when present in low copy number, could not complement the gitl-1 mutation. Second, by tetrad analysis, the
LEU2 gene on pCHY26 was unlinked to gitl, demonstrating that the plasmid had not integrated at the gitl
locus. Third, additional crosses by other git mutants
demonstrated that pCHY26 had directed integration to
the git2 locus (see Materials and methods). To rule out
the possibility that plasmid pCHY26 had integrated by
nonhomologous recombination at a site fortuitously
linked to git2, we also integrated a smaller derivative of
pCHY26, pCHY27 (Fig. 1), and demonstrated by Southern hybridization analysis and linkage analysis that it
had integrated by homologous recombination at the git2
locus (data not shown). Therefore, pCHY26 carries the
git2 gene, and git2 is a high copy number suppressor of a
gitl mutation.
Because git2 is a high copy number suppressor of a girl
mutation, we transformed strains carrying mutations in
gitl-gitl 0 {FWP113, CHP210, CHP200, CHP17, CHP75,
CHP261, CHP235, CHP60, CHP232, C H P 2 0 1 ) w i t h
pCHY26 to determine whether other git mutations
could be suppressed by git2 in high copy number. Because each strain also contained the fbpl-ura4 fusion,
suppression was assayed by monitoring the 5FOA phenotype of each transformant. Strains that carry a mutation in gitl, git2, git3, git5, git7, gitS, or gitlO were suppressed by high copy number git2 (SFOAR), whereas
transformants carrying a mutation in git4, git6, or git9
were not suppressed (5FOAS). These results suggest that
the git2 gene product may act downstream of the gene
products of gitl, git3, git5, git7, git8, and gitl O. The gene
products of git4, git6, and git9 may act either downstream of or in a separate pathway from the git2 gene
product.
The git2 gene encodes adenylate cyclase
Molecular analysis of the cloned git2 gene has shown
Downloaded from www.genesdev.org on September 10, 2007
Repression of fbpl transcription by cAMP
transcripts
B
i
BgRR
i i i
BBg
i
J
plasmid
I pCHY26
I
t
t
-]-
pCHY27
I
1
qit2 function
pCHY28
kb
Figure 1. Restriction map, transcription units, and subclones
of the git2 clone. The insert DNA present in the original git2
clone, pCHY26, and two nonfunctional subclones are shown.
The insert DNA in pCHY26 is an -10.5-kb HindIII fragment,
with additional internal HindIII sites not shown. SalI (S),
BamHI (B), BglII (Bg), and EcoRI (R) sites are marked. Function
was determined by the ability of the plasmid to transform a git2
mutant strain from 5FOAs to 5FOA R. The size, orientation, and
approximate positions of the two transcripts that hybridize to
pCHY26 were determined by a series of Northern hybridization
analyses (data not shown). The longer transcript (right) is the
git2 transcript.
that it is identical to the previously identified cyrl gene,
which encodes adenylate cyclase (Yamawaki-Kataoka et
al. 1989; Young et al. 1989). Plasmid pCHY28 (Fig. 1), a
derivative of pCHY26 in which a 2.8-kb BglII fragment
has been removed, was unable to transform git2 m u t a n t
strains to Git +. We, therefore, assumed that the git2
gene m u s t overlap some, if not all, of this portion of the
clone and initiated D N A sequencing around the leftmost BglII site (Fig. 1).
D e t e r m i n a t i o n of 340 bp of D N A sequence gave a sequence identical to that of part of the coding region for
the adenylate cyclase catalytic domain obtained by Yamawaki-Kataoka et al. (1989; EMBL/GenBank file n a m e
Yspcyrla.pl; base pairs 4084-4235 and 4261-4448) and
by Young et al. (1989; EMBL/GenBank file n a m e Yspadc.pl; base pairs 5485-5636 and 5662-5849}. Restriction
mapping of pCHY26 w i t h nine restriction endonucleases
(data not shown) was consistent with our conclusion
that git2 and cyrl are the same gene.
To verify that git2 and cyrl are the same gene and to
determine whether the G i t - phenotype can be caused by
loss of adenylate cyclase activity, we constructed git2
(cyrl) n u l l strains. In these S. pombe strains, FWP190
and FWP191, the wild-type gene is replaced by D N A that
contains a deletion of a BgllI fragment internal to git2
(see Materials and methods; Fig. 1). This deletion mutation, designated git2-1, removes part of the coding region
for the catalytic domain of adenylate cyclase and should
cause the remainder of the gene to be translated out of
frame. Northern hybridization analysis of R N A from
strain FWP191 indicates that git2-1 only alters the 5400nucleotide git2 (cyrl) transcript and has no effect on a
1600-nucleotide transcript > 700 bp from the BglII deletion (data not shown). Therefore, the git2 (cyrl) gene is
the only gene affected by the BglII deletion.
Strains that contain the git2-I n u l l allele display a
G i t - phenotype and constitutively express the fbplIacZ fusion (FWP 191; Table 1). M e m b r a n e extracts made
from strain FWP191 (git2-1) lack any detectable adenylate cyclase activity (Table 2). Therefore, loss of git2 function (adenylate cyclase) causes a G i t - phenotype. This
null allele fails to c o m p l e m e n t other git2 m u t a n t alleles,
confirming that the other git2 m u t a t i o n s are in the cyrl
gene. In contrast to S. cerevisiae, S. pombe haploid
strains that contain a null m u t a t i o n in git2 (cyrl) are
viable in the absence of cAMP. This result has also been
observed by Maeda et al. (1990) and K a w a m u k a i et al.
(1991).
cAMP represses fbpl transcription in gilt mutant and
wild-type strains
Because adenylate cyclase acts to convert ATP to cAMP,
and as m u t a t i o n s in the gene that encodes adenylate cyclase cause constitutive transcription of the fbpl gene,
we examined the effect of exogenous cAMP on transcription of the fbpl gene. Our results demonstrate that in
most git m u t a n t s and in wild-type strains, cAMP alters
fbpl transcription. The addition of 5 mM cAMP represses
transcription of the fbpl gene in a git2 m u t a n t strain as
well as in gitl, git3, git5, git7, git8, or gitl O m u t a n t s (Fig.
2). Exogenous cAMP, however, has no detectable effect
on fbpl transcription in a git6 m u t a n t (Fig. 2). Because
the git4 and git9 m u t a t i o n s do not cause a detectable
increase in fbpl transcription, we were unable to determ i n e whether there is an effect of addition of exogenous
cAMP on fbpl transcription in git4 or git9 mutants.
Expression of an fbpl-lacZ fusion in most git m u t a n t s
is also repressed by cAMP (Table 1). 13-Galactosidase assay results demonstrate that the addition of cAMP sup-
Table 1.
fbpl-lacZ
expression in g i t mutants ± cAMP
git
~-Galactosidase activity
in 8% glucosea
Strain
genotype
- cAMP
FWP70
FWP134
FWP191
FWP135
FWP136
FWP173
FWP174
FWP139
FWP140
FWP175
FWP142
FWP143
git +
gitl-I
git2-1
git2-210
git3-200
git4-17
git5-75
git6-261
git7-235
git8-60
git9-232
gitlO-201
17
1590
3486
1994
505
39
848
2396
1672
859
100
1106
±
±
+
±
+
±
±
±
+
±
±
+ cAMP
5
121
432
354
107
11
161
71
301
195
27
321
13
34
24
27
37
36
21
1674
94
20
103
20
±
±
+
±
+
±
±
±
+
±
±
+
3
3
4
5
15
3
3
114
11
3
12
2
aThe values given represent B-galactosidase sp. act. + S.E. for
two to four independent cultures. Strains were grown overnight
to 1 x 107 cells/ml in YEL (8% glucose) in the presence or
absence of 5 mM cAMP and assayed.
GENES & DEVELOPMENT
563
Downloaded from www.genesdev.org on September 10, 2007
Hoffman and Winston
Table
2. Adenylate cyclase activity and cAMP levels in
git2 mutants
Strain
allele
Adenylate
cyclase
activitya
FWP70
FWP135
FWP182
FWP183
FWP184
FWP191
git2 +
git2-210
git2-7
git2-13
git2-61
git2-!
10.4 + 0.3
8.5 _+ 1.5
8.6 + 0.8
0.3 +--0.2
1.4 _ 0.5
<0.1
git2
cAMP levelsb
repressed
3.6
2.5
5.7
0.9
2.8
0.4
-+ 0.1
+ 0.3
-+ 0.1
+--0.1
--- 0.2
+ 0.1
derepressed
3.4 -+ 0.6
ND
ND
ND
ND
ND
aThe values given represent sp. act. +- s.~. for three to four in-
dependent cultures assayed as described in Materials and methods. One unit of adenylate cyclase activity represents the production of 1 pmole of cAMP/min per mg of protein.
bcAMP levels were measured as described in Materials and
methods for cultures grown in PM medium under repressing
{8% glucose) or derepressing (0.1% glucose + 3% glycerol) conditions. Values given represent pmoles of cAMP/mg of total
protein + s.E. for two independent cultures. The low value for
cAMP measurements in the git2-1 strain was still observed after
treatment of the extract with phosphodiesterase (data not
shown). Therefore, this strain does not produce any detectable
cAMP. (ND) Not determined.
presses the constitutive expression of the fbpl-lacZ fusion in the same set of git m u t a n t s as indicated by the
Northern blot analysis of the fbpl transcript. Exogenous
cAMP has little or no effect on ~-galactosidase activity
in git4, git6, or git9 m u t a n t strains.
Addition of exogenous cAMP also represses fbpl transcription in both git2 + and git2-210 strains grown under
derepressing conditions (Fig. 3). Similarly, cAMP addition represses expression of an fbpl-IacZ fusion in a git +
strain (FWP70) grown under derepressing conditions:
This strain contained 1549 + 47 units of ~-galactosidase
activity {sp. act. -+ S.E. for four independent liquid cultures) in the absence of cAMP and only 122 _+ 25 units in
the presence of cAMP. In addition, cAMP caused cell
death w h e n FWP70 was grown in the presence of 0.1%
glucose and 3% glycerol (data not shown), probably due
to the inability of the cells to use glycerol as a carbon
source owing to the simulation of glucose repression by
cAMP. These results demonstrate that exogenous cAMP
is sufficient for repression of transcription from the fbpl
promoter in a wild-type strain grown under normally
derepressing conditions.
git2 intragenic complementation
We determined previously that diploids formed from
strains carrying the git2-7 and git2-61 alleles have a Git +
phenotype, indicating intragenic complementation
(Hoffman and Winston 1990). Therefore, strains that
contain the git2-7 and git2-61 alleles have been tested for
their ability to c o m p l e m e n t strains carrying 31 independent git2 mutations. Diploids carrying either the git2-7
or git2-61 allele on one homolog and each of the 31 git2
alleles on the other homolog were tested for complemen-
564
GENES & DEVELOPMENT
tation as described in Materials and methods. We also
tested all combinations of the 17 original h - git2 isolates
by the 14 original h + git2 isolates for intragenic complementation.
On the basis of these c o m p l e m e n t a t i o n tests, 11 of the
31 git2 mutants display at least some intragenic complementation. Eight m e m b e r s form one group, and three
members form the second group. The remaining 20 spontaneous git2 m u t a n t alleles, as well as the git2-1 null
allele, do not c o m p l e m e n t any git2 allele. The first complementation group consists of three strongly complem e n t i n g alleles, git2- 7, git2-110, and git2-266, four moderately c o m p l e m e n t i n g alleles, git2-18, git2-68, git2-216,
and git2-226, and one weakly c o m p l e m e n t i n g allele,
git2-210. The second c o m p l e m e n t a t i o n group consists of
the strongly c o m p l e m e n t i n g allele git2-61 and two moderately c o m p l e m e n t i n g alleles, git2-29 and git2-83. An
example of the c o m p l e m e n t a t i o n analysis is shown in
Figure 4. Because most spontaneous git2 m u t a n t alleles
do not c o m p l e m e n t alleles of either c o m p l e m e n t a t i o n
group, these groups represent intragenically complem e n t i n g alleles and not m u t a t i o n s in two tightly linked
genes.
Adenylate cyclase activity and cAMP levels in git2
mutant strains
To determine whether all git2 m u t a t i o n s cause a loss or
reduction of adenylate cyclase activity, adenylate cyclase activity was assayed in m e m b r a n e extracts from
git + and a set of git2 m u t a n t strains. These assays were
performed in the presence of M n ~ +, as in vitro activation
of S. pombe adenylate cyclase in the presence of Mg 2+
has not been observed (Yamawaki-Kataoka et al. 1989;
Engelberg et al. 1990; C.S. Hoffman and F. Winston, unpubl.I. The results of the assays (Table 2) show that the
git2-61 and git2-13 m u t a t i o n s significantly reduce adenylate cyclase activity, whereas the git2-1 null allele
causes the complete loss of adenylate cyclase activity.
The git2-7 and git2-210 mutations, however, have little
or no effect on adenylate cyclase activity as measured in
vitro. Although the four spontaneous git2 mutations
have different effects on adenylate cyclase activity, they
all increase [bpl-iacZ expression to a similar high constitutive level [data not shown). Therefore, git2 alleles
that confer very different levels of adenylate cyclase activity as measured in vitro all corder a similar G i t - phenotype in vivo.
To further analyze the different classes of git2 mutants, we also measured steady-state cAMP levels in
wild-type and git2 m u t a n t strains (Table 2). The git2-1
(null) and git2-13 m u t a n t s both have greatly reduced levels of cAMP. (The low value that we obtained for cAMP
m e a s u r e m e n t s in the git2-1 strain was still observed
even after treatment of the extract w i t h phosphodiesterase (data not shown). Therefore, this strain does not
produce any detectable cAMP.) The other git2 mutants
tested do not have significantly reduced cAMP levels. In
addition, steady-state cAMP levels in a git2 + strain are
similar for growth under conditions repressing or dere-
Downloaded from www.genesdev.org on September 10, 2007
Repression
git +
cAMP
--
+
gitl-1
I~' '
--
+
git2.210
II
--
+
git3.200
II
--
"
+
git4-17
II
--
+
git5-75
II
--
+
git6-261
II
--
+
git8-60
git7-235
II
--
+
II
-
+
of tbpl
transcription
by cAMP
git9.232 gitlO,201
II
--
+
II
-
=:= I
, :+:
fbp l
leul -4~
2. Exogenous cAMP represses fbpl transcription in most git mutant strains. Northern hybridization analysis was performed
on RNA from a git + strain (FWP70) and from strains carrying mutations in gitl-gitlO (FWP134, FWP135, FWP136, FWP173, FWP174,
FWP139, FWP140, FWP175, FWP142, and FWP143; see Table 1 for corresponding fbpl-lacZ expression), grown under repressing
conditions (8% glucose) in the presence and absence of 5 mM cAMP. The filter was hybridized to 32p-labeled plasmids pAV06 (to detect
fbpl RNA; Vassarotti and Friesen 1985) and pYK311 (to detect leul RNA as an internal standard; Kikuchi et al. 1988), and exposed
Figure
to X-ray film.
pressing for fbpl expression. Therefore, although some
git2 alleles abolish or reduce cAMP levels and cause a
G i t - phenotype, some other git2 alleles cause a G i t phenotype, although they cause little change in steadystate cAMP levels.
fbpl transcription is controlled by a cAMP-dependent
protein kinase
In other eukaryotic cAMP signaling pathways, the role
of cAMP is to activate cAMP-dependent protein kinase (cAPK), which then phosphorylates target proteins
(Beebe and Corbin 1986). To determine whether the role
of cAMP in S. pombe glucose repression is to activate
cAPK, we examined fbpl transcription in a strain containing a mutation in the cgsl gene that encodes the
regulatory subunit of cAPK (M. McLeod, pers. comm.).
On the basis of constitutive transcription of fbpl in a
strain lacking adenylate cyclase activity, if cgsl is in the
same pathway, a cgsl mutation would be expected to
reduce or eliminate fbpl expression in derepressed cells.
For this experiment we used a mutant that contains the
spontaneous cgsl-1 mutation (M. McLeod, pers. comm.).
The results demonstrate that cgsl-1 reduces fbpl transcription in a strain grown under derepressing conditions
(0.1% glucose and 3% glycerol; Fig. 5). Furthermore, the
cgsl-1 mutation greatly reduces fbpl-lacZ expression in
strains grown under derepressing conditions (3% maltose; Table 3). These results strongly suggest that cAMP
represses fbpl transcription by the activation of cAPK.
The more dramatic effect of the cgsl mutation observed
in Table 3, relative to Figure 5, could be due to the fact
that 3% maltose is a less derepressing carbon source
than 0.1% glucose plus 3% glycerol.
To characterize the role of cgsl in this pathway in
greater detail, we carried out epistasis tests with cgsl
and git2, and with cgsl and git6. As expected, the cgsl-1
git2-1 double mutant resembled the cgsl-1 m u t a n t (Table 3), that is, the cgsl-1 mutation suppressed the constitutive expression of the fbpl-lacZ fusion conferred by
the git2-1 allele. This result supports the idea that glucose repression of fbpl is mediated by cAPK. Interestingly, the cgsl-1 git6-261 double mutant is constitutive
for expression of fbpl-lacZ, consistent with the observation that the git6-261 mutation is not suppressed by
exogenous cAMP. This epistasis test suggests that the
. . . . ~~ :~i~:~/~)i~)~ii~:i~:i~(//~i!~~i~
-5FOA
+5FOA
day 3
day 3
+5FOA
day 7
4. Certain git2 mutant alleles display strong intragenic
complementation. Complementation analyses were performed
as described in Materials and methods. Two diploids from each
cross were colony-purified and replica-plated to media with and
without 5FOA at 30°C. Diploids that contain a git2 + allele are
5FOA R, demonstrating that the git2 mutations are recessive.
Diploids that contain two git2 mutations are usually 5FOA s
(noncomplementation), except those that contain git2-7 with
git2-61, which are 5FOA R {growth on 5FOA by day 3; strong
complementation) and those that contain git2-210 with git2-61,
which are weakly 5FOA R {growth on 5FOA by day 7; weak
complementation). Rare, individual 5FOA R colonies that appear
in 5FOA s patches likely represent events other than complementation, such as mitotic recombination at git2, second site
suppressor mutations, or reversion events.
Figure
-
+
-
+
•-
+
.~.
÷
Figure
3. Exogenous cAMP represses fbpl transcription in a
git2 + and a git2- mutant strain grown on a derepressing carbon
source. Northern hybridization analysis was performed on RNA
from FWP70 (git2 +) and from FWP135 (git2-210) grown on a
repressing carbon source (8% glucose) and a derepressing carbon
source (3% maltose), in the presence and absence of 5 mM
cAMP, and probed as described in Fig. 2.
GENES & DEVELOPMENT
565
Downloaded from www.genesdev.org on September 10, 2007
Hoffman and Winston
wild type
cgs l-1
ras l "
~ vatlZ
ras 7
' R D "i R D I[ R D if'R: D
':
Figure 5. fbpl transcription is reduced in a cgsl mutant strain
and is normal in strains carrying rasl mutations. Northern hybridization analysis was performed on RNA from strains
FWP180 (cgsl + rasl +), FWP177 (cgsl-1), FWP179 (rasl-), and
FWP181 (rasl"an7; activated form) grown under repressing (8%
glucose) and derepressing {0.1% glucose and 3% glycerol) conditions. Additional bands present in lanes containing FWP180,
FWP179, and FWP181 RNAs are probably due to transcription
of the LEU2 gene integrated adjacent to rasl.
git6 gene product does not normally participate in the
stimulation of adenylate cyclase but, rather, acts more
directly in repression of fbpl transcription.
Mutations in rasl do not alter fbpl transcription
S. cerevisiae adenylate cyclase is activated by the products of the RAS1 and RAS2 genes (Toda et al. 1985). In S.
pombe, however, experiments indicate that the rasl
gene does not carry out a similar function (Fukui et al.
1986; Nadin-Davis et al. 1986b, 1989). To investigate
further the role of rasl with respect to adenylate cyclase
activity, we examined fbpl transcription in strains that
contain either a rasl - or a rasl van7 (activated) mutation
(Nadin-Davis et al. 1986b). Northern hybridization analysis {Fig. 5) demonstrates that there is wild-type regulation of fbpl transcription in both strains. This result
demonstrates further that S. pombe rasl does not regulate adenylate cyclase activity.
Discussion
Our previous work identified a set of genes required for
glucose repression of fbpl transcription in S. pombe
Table 3. f b p l - l a c Z expression in cgsl-1, cgsl-1 git2-1 and
¢gsl-1 git6-261 mutants
~-Galactosidase a c t i v i t y ~
Strain
Relevant
genotype
8% glucose
3% m a l t o s e
FWP189
FWP191
FWP139
FWP192
FWP193
FWP194
cgsl + git +
git2-1
git6-261
cgsl-I
cgsl-1 git2-1
cgsl-1 git6-261
12
3486
2396
21
18
3417
1508
5444
7400
118
116
4621
-+ 0
+ 432
+ 71
+ 6
- 5
+ 264
+-- 212
-+ 1043
+ 2413
+ 8
+ 14
+ 1078
aThe values given represent [3-galactosidase sp. act. -+ S.E. for
two or three independent cultures. Strains were grown overnight in YEL to 1 x 107 cells/ml, with either 8% glucose (repressing) or 3% maltose (derepressing) as the carbon source, and
assayed.
566
GENES & DEVELOPMENT
(Hoffman and Winston 1990). Our present work demonstrates that one of these genes, git2, is the same as the
previously identified cyrl gene, encoding adenylate cyclase, and that cells that lack this activity are unable to
repress fbpl transcription. Our experiments also demonstrate that exogenous cAMP represses fbpl transcription
in both wild-type and many, but not all, git mutant
strains. These results strongly suggest that S. pombe responds to glucose by activating adenylate cyclase as part
of a cAMP signaling pathway.
The adenylate cyclase in vitro assays performed here
almost certainly represent the unstimulated activity of
adenylate cyclase and do not reflect the ability of adenylate cyclase to respond to in vivo stimuli. Adenylate
cyclase activity from S. pombe is not stimulated by the
addition of guanine nucleotides in vitro, and activity has
only been observed when assayed in the presence of
Mn2+; activity in the presence of Mg 2+ has not been
detected (Yamawaki-Kataoka et al. 1989; Engelberg et al.
1990; C.S. Hoffman and F. Winston, unpubl.). Our results suggest that the addition of glucose may cause a
post-translational modification of adenylate cyclase that
stimulates activity in vivo but that is not detectable by
current in vitro assays. Clearly, an in vitro assay that
reflects more clearly in vivo activity and stimulation
will be necessary to allow a more accurate assessment of
the adenylate cyclase defects in the different git2 mutants.
Differences in the adenylate cyclase levels among git2
mutants and the observation of git2 intragenic complementation suggest that S. pombe adenylate cyclase contains separable domains and that it functions in a complex containing two or more copies of the git2 (cyrl) gene
product. Genetic evidence for such a complex in S. cerevisiae has been obtained by Field et al. (1990b). That
mutations in each git2 intragenic complementation
group represent alterations of different functional domains of adenylate cyclase is consistent with results of
adenylate cyclase assays for different git2 mutants (Table
2). Two strains that contain mutations in the first complementation group (git2-7 and git2-210) possess virtually wild-type levels of adenylate cyclase activity as measured in vitro and are therefore unlikely to affect directly
the adenylate cyclase catalytic domain. In contrast, a
strain that contains a mutation in the second complementation group (git2-61) possesses significantly reduced adenylate cyclase activity as does a strain that
contains a mutation that fails to complement members
of either group (git2-13).
The cAMP levels in some git2 mutants are reduced
relative to that of a wild-type strain (Table 2), supporting
the hypothesis that these mutations cause a reduction in
the ability of the cell to produce cAMP in response to
glucose. The facts that a git2 null mutant, which makes
no detectable cAMP, causes derepression of fbpl transcription, and that addition of exogenous cAMP restores
repression, demonstrate that cAMP causes fbpl repression. The result that some git2 mutants do not have
reduced cAMP levels may reflect feedback control of
cAMP levels as in S. cerevisiae (Nikawa et al. 1987). In
Downloaded from www.genesdev.org on September 10, 2007
Repression of [bpl transcription by cAMP
addition, wild-type cultures grown under repressing and
derepressing conditions possess similar steady-state levels of cAMP. These results suggest that the steady-state
level of cAMP is not altered by glucose repression.
Rather, the cAMP signal is likely to be transient in cells
exposed to glucose, as is the case in S. cerevisiae (Mbonyi
et al. 1988; Thevelein 1988). Measurement of cAMP levels i m m e d i a t e l y after glucose addition in wild-type and
git strains will test whether cAMP levels are altered in
response to glucose.
These studies provide us with likely candidates for
genes required for the activation of adenylate cyclase in
S. pombe. These candidates include the genes gitl, git3,
gitS, git7, git8, and gitlO, as the G i t - phenotype caused
by mutations in these genes is suppressed by exogenous
cAMP and by the git2 gene in high copy number. One or
more of these git genes m a y encode subunits of a trimeric guanine nucleotide-binding protein (G protein)
that stimulates adenylate cyclase activity, as has been
observed for m a m m a l i a n cells (Gilman 1984). Because
the G protein m a y have to assemble and/or couple to the
glucose receptor before stimulation of adenylate cyclase,
mutations in any of the genes encoding G protein subunits could produce a G i t - phenotype. One or more git
genes m a y also encode a glucose receptor, proteins that
modulate G protein activity [analogous to CDC25 of S.
cerevisiae (Broek et al. 1987; Robinson et al. 1987)], or
additional subunits of the active adenylate cyclase complex [analogous to the 70-kD adenylate cyclase-associated protein of S. cerevisiae (Fedor-Chaiken et al. 1990;
Field et al. 1990a)]. Alternatively, stimulation of adenylate cyclase in S. pombe m a y occur by a different type of
pathway.
Glucose repression of fbpl in S. pombe m a y be analogous in some ways to glucose repression of ADH2 in S.
cerevisiae. In S. cerevisiae, studies have demonstrated
that glucose induces a transient cAMP signal and that
this response requires a functional RAS1 or RAS2 gene
(Mbonyi et al. 1988; Thevelein 1988). The cAMP signal
activates cAPK; for ADH2, glucose repression occurs via
this pathway, as cAPK appears to phosphorylate and inactivate the ADH2 transcriptional activator ADR1
(Cherry et al. 1989; Taylor and Young 1990). Cherry et al.
(1989) have also shown that mutations in BCY1 (encoding the regulatory subunit of cAPK) reduce expression of
ADH2. The result that a mutation in S. pombe cgsl (encoding the regulatory subunit of cAPK) reduces fbpl
derepression indicates that the role of cAMP in glucose
repression of fbpl is to stimulate cAPK. (Interestingly,
the ability of a git2-1 cgsl-1 strain to show m i l d derepression of fbpl-lacZ expression (Table 3) suggests that
glucose also causes repression of fbpl transcription by a
cAMP-independent mechanism.)
From our results thus far, we cannot conclude that
cAPK acts to repress fbpl transcription by the same
m e c h a n i s m as seen for ADH2 in S. cerevisiae. Mutations
in git6 are not suppressed by exogenous cAMP or by a
m u t a t i o n in cgsl (Fig. 3; Table 3). Possibly, git6 encodes
the catalytic subunit of cAPK that inactivates the transcriptional activator of fbpl. However, the git6 gene
could also encode a transcriptional repressor that is activated by cAPK. The cAMP-responsive element-binding
protein (CREB} from rat can act as either a transcriptional activator or a repressor, depending on its phosphorylation state, at the c-jun promoter (Lamph et al. 1990).
Previous work has shown that transcription of the
mei2 gene of S. pombe is also repressed by cAMP (Watanabe et al. 1988). This gene is required for entry into
meiosis and is inducible by nitrogen starvation; however, the role of cAMP in meiosis is still unclear (for
review, see McLeod 1989). Therefore, S. pombe m a y regulate cAMP levels w i t h respect to both the carbon source
and the presence of a nitrogen source. It is likely that
additional factors are involved that distinguish between
response to glucose and response to a nitrogen source.
Our future work will address how adenylate cyclase is
activated in S. pombe and how cAPK acts to repress fbpl
transcription. By continued molecular and genetic analyses of git genes, and by analysis of the fbpl promoter,
we hope to further our understanding of the role of signal
transduction and transcriptional repression of fbpl expression.
Materials and m e t h o d s
Yeast strains and media
All S. pombe strains used in this study are listed in Table 4.
Genetic nomenclature of S. pombe follows rules proposed by
Kohli (1987). The fbpl::ura4 allele is a translational fusion that
disrupts the fbpl gene, and the ura4::fbpl-lacZ allele' is a disruption of the ura4 gene by the fbpl-lacZ translational fusion
(Hoffman and Winston 1990). Standard rich media YEA and YEL
(Gutz et al. 1974) were supplemented with 2% casamino acids.
Minimal media (SD) supplemented with amino acids and synthetic complete media (SC) lacking a specific amino acid (Sherman et al. 1978) were used with modifications described previously (Hoffman and Winston 1990). PM medium (defined minimal medium; Watanabe et al. 1988) was supplemented with
required amino acids. Carbon sources were generally present at
a concentration of 3%, and strains were grown at 30°C, unless
otherwise specified. Sensitivity to 5FOA (PCR, Inc.) was determined on SC, SC-leu, and SC-ade solid media containing 8%
glucose, 50 rag/liter of uracil, and 0.4 g/liter of 5FOA. When
indicated, cAMP (Aldrich Chemical Company, Inc.) was added
to a final concentration of 5 mM. Crosses were done on either
MEA (Gutz et al. 1974) containing 0.4% glucose or on YPD
(Sherman et al. 1978).
Recombinant DNA methodology
Standard recombinant DNA techniques (including DNA restriction digests, ligations, and bacterial transformations) were
done according to Maniatis et al. (1982). E. coli strains HB101
(Boyer and Roulland-Dussoix 1969) and JM101 (Messing 1979)
were the host strains for bacterial transformations. Yeast transformations were done by the lithium acetate method (Ito et al.
1983). Small-scale plasmid preparations from E. coli were done
by the alkaline lysis method (Bimboim and Doly 1979). Smallscale DNA preparations from yeast to recover plasmids in E.
coli were done as described previously (Hoffman and Winston
1987). DNA fragment isolation was done by electroelution. Restriction endonucleases SalI and XhoI were purchased from Boe-
GENES & DEVELOPMENT
567
Downloaded from www.genesdev.org on September 10, 2007
Hoffman and Winston
Table 4.
Strains
Strain
Genotype
CHP1
FWP148
FWP167
I:WP101
FWPll2
FWP113
CHP210
FWP 189
FWP190
CHP200
CHP 17
CHP75
CHP261
GHP235
CHP60
CHP232
CHP201
FWP70
FWP134
FWP135
FWP136
FWP 173
FWPI74
FWP139
FWP140
FWP175
FWP142
FWP143
FWPI82
FWP 183
FWP184
FWP191
CHP7
CHP13
CHP61
FWPll4
FWP169
FWP177
FWP 180
FWP179
FWP181
FWP192
FWP193
FWP194
h ade6-M216 lys1-131 leul-32 ura4::fbpl-lacZ fbpl ::ura4 gitl-1
h + lysl-13i leul-32 ura4: :fbpl-lacZ fbp l ::ura4 [pCHY26]
h + ade6--M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 [pCHY27]
h + ade6-M210 his7-366 Ieul-32 ura4::fbpl-IacZ fbpl ::ura4
h - ade6--M216 his7-366 leul-32 ura4 ::fbp l-lacZ fbp l ::ura4
h + ade6--M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 git1-1
h - ade6-M216 his7-366 leul-32 ur a4 ::fbp l-lacZ fbp l ::ura4 git2-210
h + ade6--M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 git2 + int::LE U2
h + ade6-M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git2-1 int::LE U2
h - ade6-M216 his7-366 leul-32 ura4 ::fbp l-lacZ fbp l ::ura4 git3-200
h + ade6--M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git4-17
h + ade6--M210 his7-366 leul-32 ura4 ::fbp l-lacZ fbp l ::ura4 git5- 75
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 git6-261
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git7-235
h + ade6-M210 his 7-366 leul-32 ura4: :fbp l-lacZ fbp l ::ura4 git8-60
h - ade6-M216 his7-366 leul-32 ur a4 ::fbp l-lacZ fbp l ::ura4 git9-232
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 gitl 0-201
h + ade6-M216 lys1-131 leul-32 ura4::fbpl-lacZ
h + ade6-M210 leul-32 ura4::fbpl-lacZ girl-1
h - his7-366 leul-32 ura4::fbpl-lacZ git2-210
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ git3-200
h - leul-32 ura4::fbpl-lacZ git4-17
h - leul-32 ura4::fbpl-tacZ git5-75
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ git6-261
h - leul-32 ura4::fbpl-lacZ git7-235
h - leul-32 ura4::fbpl-lacZ git8-60
h + leul-32 ura4::fbpl-lacZ git9-232
h + leul-32 ura4::fbpl-lacZ gitlO-201
h - leul-32 ura4::fbpl-lacZ git2-7
h - leuI-32 ura4::fbpl-lacZ git2-13
h - leul-32 ura4::fbpl-lacZ git2-61
h - ade6--M210 leul-32 ura4::fbpl-lacZ git2-1 int::LEU2
h + ade6--M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 git2- 7
h + ade6-M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git2-13
h + ade6-M210 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git2-6l
h - ade6--M216 his7-366 leul-32 ura4::fbpl-lacZ fbpl ::ura4 git2-7
h - ade6-M216 his7-366 leul-32 ura4::fbpl-lacZ fbpl::ura4 git2-61
h 9° ade6--M210 leul-32 cgsl-I
h + ade6-M216 leul-32 rasl + int::LEU2BglH
h + ade6-M216 leul-32 ura4- rasl - int::LEU2HincH
h 9° ade6-M210 leul-32 rasl vallz int::LEU2BglH
h 9° ade6--M216 leul-32 ura4::fbpl-lacZ cgsl-1
h + ade6--M216 leul-32 his7-366 ura4::fbpl-lacZ cgsl-1 git2-1 int::LEU2
h + ade6--M216 leul-32 his7-366 ura4::fbpl-lacZ cgsl-1 git6-261
+
hringer Mannheim Biochemicals. T4 DNA polymerase, the
large fragment of DNA polymerase I (Klenow), DNA primers,
and all other restriction enzymes were purchased from New
England Biolabs.
Cloning of the git2 gene
The git2 gene was cloned as a high copy number suppressor of
the gitl-1 mutation in strain CHP1 (Table 4). CHP1 was transformed to Leu + on SC-leu (8% glucose) with an S. pombe genomic library (partial HindIII-digested DNA cloned into the
HindIII site of pWH5; Wright et al. 1986; P. Young and D. Beach,
unpubl.), and transformants were replica-plated to SC-leu and
SC-leu with 5FOA to screen for plasmids that restored regulation of the fbpl-ura4 fusion. Plasmids were isolated from 10
568
GENES & DEVELOPMENT
5FOA R transformants and used to transform E. coli strain
HB101 to Amp R. Plasmid preparations from the E. coli transformants were used for restriction analyses and retransformation
of CHP1. A single candidate plasmid, pCHY26 {Fig. 1), conferred
a strong 5FOAR phenotype on retransformation of CHP1. Other
candidate plasmids conferred a weak 5FOAR phenotype and
have not been examined further.
Plasmid pCHY26 was integrated into the S. pombe genome by
transformation of CHP1 with PmlI-digested pCHY26. Analysis
of transformants indicated that pCHY26 did not carry the gitl
gene, as originally anticipated. In a cross, the G i t - phenotype
associated with the gitl-1 mutation and the Leu + phenotype
associated with the LEU2 gene from pCHY26 segregated independently in tetrads. Additional genetic crosses with a progeny
from this cross, FWP148 (Git + Leu+), however, demonstrated
Downloaded from www.genesdev.org on September 10, 2007
Repression of fbpl transcription by cAMP
that these phenotypes did cosegregate (nine of nine PD tetrads)
when crossed with the git2 mutant CHP210. Although this genetic analysis indicated that pCHY26 carried the git2 gene, we
were unable to determine whether pCHY26 had integrated by
homologous recombination (as judged by Southern hybridization analysis) owing to the large size of the insert DNA in this
plasmid. To determine whether the plasmid insert directs homologous integration at git2, we constructed plasmid pCHY27
{Fig. 1) by digesting pCHY26 with BamHI and recircularizing
the large restriction fragment. Plasmid pCHY27 contains most
of the vector DNA plus - 4 kb of S. pombe insert DNA. Homologous integration into strain FWP101 (Git + Leu-), creating
strain FWP 167 (Git + Leu + ), was confirmed by Southern hybridization analysis (Southern 1975) of both EcoRI and EcoRV digests of transformant DNA. In a cross of FWP167 with strain
CHP210, the Git + and Leu + phenotypes cosegregated (25 PD
tetrads and one TT tetrad). Therefore, pCHY27 contains genomic DNA from the git2 locus.
whereas diploids formed between members of the same groups
were 5FOA s (noncomplementation).
~-Galactosidase assays and Northern hybridization analysis
[3-Galactosidase assays and Northern hybridization analysis
were done as described previously (Hoffman and Winston 1990).
Plasmid pAV06 (Vassarotti and Friesen 1985) was used as a
probe for the fbpl transcript. RNA amounts were standardized
by hybridization to plasmid pYK311, which carries the leul
gene (Kikuchi et al. 1988).
cAMP assays
Construction of the git2-1 null allele
Extracts for cAMP assays were prepared from cultures grown to
1 x 107 cells/ml in PM medium according to the method of
Fedor-Chaiken et al. (1990) and assayed using the radioimmune
assay kit available from Amersham. cAMP levels were standardized to total protein concentration using the BCA protein
assay (Smith et al. 1985). Total protein extracts were made in
0.2 N NaOH by vortexing cells in the presence of glass beads.
A git2 null allele was constructed by digesting pCHY26 with
BglII and recircularizing the large BglII fragment, resulting in
Adenylate cyclase assays
plasmid pCHY28 (Fig. 1). This plasmid is unable to complement
mutations in git2. On the basis of DNA sequence analysis of the
git2 gene (cyrl; Yamawaki-Kataoka et al. 1989; Young et al.
1990} this BglII deletion removes part of git2 that encodes the
adenylate cyclase catalytic domain and shifts the remainder of
the coding region out of frame with the amino-terminal coding
region. Plasmid pCHY28 was digested with ClaI, followed by
ligation of the large restriction fragment, resulting in removal of
much of the vector DNA. This plasmid, pCHY29, contains a
single SalI site, present in the insert DNA. pCHY29 was then
digested with SalI and ligated with a 2.2-kb SalI-XhoI fragment
from YEpl3 (Broach et al. 1979), which carries the LEU2 gene of
S. cerevisiae, to form plasmid pCHY30. (The LEU2 gene of S.
cerevisiae complements the leul-32 mutation in S. pombe.)
Strain FWP101 was transformed to Leu + with a HindIII fragment from pCHY30 that carries the git2 BglII deletion (git2-1)
and the LEU2 insertion. One transformant (FWP190) was confirmed to contain a single copy of the git2-1 null allele and the
adjacent LE U2 insertion by Southern hybridization analysis. In
a cross between FWP 190 (Git- Leu +) and FWP 112 (Git + Leu-),
the Git- phenotype cosegregates with the Leu* phenotype. The
git2-1 mutation is also unable to complement other git2 mutations (Fig. 4). Strain FWP189, which contains the LEU2 insertion, but adjacent to git2 +, is Git + . In addition, integration of
the LEU2 gene adjacent to the git2 + gene has no effect on expression of an fbpl-lacZ fusion (Table 3; strain FWP189); therefore, it is the git2-1 deletion and not the LEU2 insertion that
causes the G i t - phenotype.
Adenylate cyclase activity was assayed in membrane extracts as
described by Casperson et al. (1985) with some modifications.
Strains were grown in 250 ml of YEL (8% glucose) to a density
of - 1 x 10z cells/ml. The cultures were poured over an equal
volume of ice and subsequently maintained at 0°C to optimize
recovery of active adenylate cyclase. Cells were pelleted and
washed twice, once with 30 ml water and once with 10 ml of
buffer A [50 mM Mes (pH 6.2), 1 mM MgCI~, 1 mM EGTA, 1 mM
~-mercaptoethanol], and resuspended in 0.4 ml of buffer A with
1 mM PMSF. Glass beads were added to the meniscus, and cells
were lysed by vortex-mixing for 4 min (done in 30-sec intervals
followed by at least 1 rain on ice). Lysates were transferred to
Eppendorf tubes (glass beads were washed with 1 ml of buffer A
and pooled with the lysatesJ and centrifuged for 15 min in a
microfuge. Membrane extracts were gently collected by resuspending the transluscent section of the pellet in 100 ~1 of buffer
A with 10% glycerol and were stored at - 70°C.
Each reaction mix contained 20-25 ~g of membrane extracts
in 100 mM Mes (pH 6.2), 5 mM MnC12, 0.1 mg/ml of BSA, 20 mM
creatine phosphate, 20 units/ml of creatine phosphokinase, 0.1
mM EGTA, 2 mM ~-mercaptoethanol, 1.3 mM [cxa2p]ATP {20-65
cpm/pmole), and 0.5 mM 3H-labeled cAMP {5000 cpm 1. Each
reaction was initiated by the addition of the membrane extract,
proceeded for 30 min at 30°C, and was terminated by the addition of 1 ml of 0.2% SDS, 0.1 mM cAMP, and 1.2 mM ATP,
followed by boiling for 3 min. The amount of 32P-labeled cAMP
produced was determined by the procedure of Salomon et al.
{1974). Assays were linear with respect to time and membrane
extract concentration under these conditions.
Complementation analysis
Complementation analyses were carried out by determining the
sensitivity or resistance to 5FOA of diploids formed between
git2 mutant strains, as described previously (Hoffman and Winston 1990). Complementation between two recessive git2 alleles restores regulated expression of the fbpl-ura4 fusion, resulting in a 5FOAR phenotype when grown under repressing conditions (8% glucose). Noncomplementation results in constitutive expression of the fbpl-ura4 fusion, resulting in a
5FOAs phenotype. Diploids formed between any two members
of the two groups that displayed git2 intragenic complementation showed partial to complete 5FOAR (complementation),
Acknowledgments
We thank Maureen McLeod for sharing unpublished data and
for providing strains carrying the cgsl-1 and rasl mutations,
Masayuki Yamamoto for sharing unpublished data, and David
Beach for the S. pombe genomic library used to clone the git2
gene. We thank Scott Cameron, Eva Neer, Ken Ferguson, Yuriko
Yamawaki-Kataoka, and Hye-Ryun Choe for helpful discussions regarding adenylate cyclase activity and in vitro assays
during the course of these studies. We also thank Greg Prelich,
Karen Amdt, and Elizabeth Malone for critical reading of this
manuscript, and David Eisenmann and Greg Prelich for assis-
GENES ,9,,DEVELOPMENT
569
Downloaded from www.genesdev.org on September 10, 2007
Hoffman and Winston
tance in preparation of the figures. This work was supported by
American Cancer Society postdoctoral grant PF-2853 to C.S.H.,
and National Science Foundation grant DCB8451649 and a
grant from the Stroh Brewery Company to F.W.
The publication costs of this article were defrayed in part by
payment of page charges. This article must therefore be hereby
marked "advertisement" in accordance with 18 USC section
1734 solely to indicate this fact.
References
Beebe, S.J. and J.D. Corbin. 1986. Cyclic nucleotide protein kinases. In The enzymes (ed. P.D. Boyer), vol. 17, pp. 43-111.
Academic Press, New York.
Bimboim, H.C. and J. Doly. 1979. A rapid alkaline extraction
procedure for screening recombinant plasmid DNA. Nucleic
Acids Res. 7: 1513-1523.
Boeke, J.D., F. Lacroute, and G.R. Fink. 1984. A positive selection for mutants lacking orotidine-5'-phosphate decarboxylase activity in yeast: 5-Fluoro-orotic acid resistance. Mol.
Gen. Genet. 197: 345-346.
Botsford, J.L. 1981. Cyclic nucleotides in procaryotes. Microbiol. Rev. 45: 620-642.
Boyer, H.W. and D. Roulland-Dussoix. 1969. A complementation analysis of the restriction and modification of DNA in
E. coli. 1. Mol. Biol. 41: 458-472.
Broach, J.R., J.N. Strathem, and J.B. Hicks. 1979. Transformation in yeast: Development of a hybrid cloning vector and
isolation of the CANI gene. Gene 8: 121-133.
Broek, D., N. Samiy, O. Fasano, A. Fujiyama, F. Tamanoi, J.
Northup, and M. Wigler. 1985. Differential activation of
yeast adenylate cyclase by wild-type and mutant ras proteins. Cell 41: 763-769.
Brock, D., T. Toda, T. Micheali, L. Levin, C. Birchmeier, M.
Zoller, S. Powers, and M. Wigler. 1987. The S. cerevisiae
CDC25 gene product regulates the RAS/adenylate cyclase
pathway. Cell 48: 789-799.
Casperson, G.F., N. Walker, and H.R. Bourne. 1985. Isolation of
the gene encoding adenylate cyclase in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. 82: 5060-5063.
Cherry, J.R., T.R. Johnson, C. Dollard, J.R. Shuster, and C.L.
Denis. 1989. Cyclic AMP-dependent protein kinase phosphorylates and inactivates the yeast transcriptional activator
ADR1. Ceil 56: 409-419.
Edelman, A.M., D.K. Blumenthal, and E.G. Krebs. 1987. Protein
serine/threonine kinases. Annu. Rev. Biochem. 56: 567-613.
Engelberg, D., E. Poradosu, G. Simchem, and A. Levitzki. 1990.
Adenylyl cyclase activity of the fission yeast Schizosaccharomyces pore be is not regulated by guanyl nucleotides. FEBS
Lett. 261: 413--41K
Eraso, P. and J.M. Gancedo. 1984. Catabolite repression in yeast
is not associated with low levels of cAMP. Eur. J. Biochem.
141: 195-198.
Fedor-Chaiken, M., R.J. Deschenes, and J.R. Broach. 1990.
SRV2, a gene required for RAS activation of adenylate cyclase in yeast. Cell 61: 329-340.
Field, J., A. Vojtek, R. Ballester, G. Bolger, J. Colicelli, K. Ferguson, J. Gerst, T. Kataoka, T. Michaeli, S. Powers, M. Riggs,
L. Rodgers, I. Wieland, B. Wheland, and M. Wigler. 1990a.
Cloning and characterization of CAP, the S. cerevisiae gene
encoding the 70 kd adenylyl cyclase-associated protein. Cell
61: 319-327.
Field, J., H.-P. Xu, T. Michaeli, R. Ballester, P. Sass, M. Wigler,
and J. Colicelli. 1990b. Mutations of the adenylyl cyclase
gene that block RAS function in Saccharomyces cerevisiae.
570
GENES & DEVELOPMENT
Science 247: 464-467.
Fukui, Y. and Y. Kaziro. 1985. Molecular cloning and sequence
analysis of a ras gene from Schizosaccharomyces pombe.
EMBO ]. 4: 687-691.
Fukui, Y., T. Kozasa, Y. Kaziro, T. Takeda, and M. Yamamoto.
1986. Role of a ras homolog in the life cycle of Schizosaccharomyces pombe. Cell 44: 329-336.
Gancedo, J.M. and C. Gancedo. 1986. Catabolite repression mutants of yeast. FEMS Microbiol. Rev. 32: 179-187.
Gilman, A.G. 1984. G proteins and dual control of adenylate
cyclase. Cell 36: 577-579.
Gutz, H., H. Heslot, U. Leupold, and N. Loprieno. 1974.
Schizosaccharomyces pombe. In Handbook of Genetics (ed.
R.C. King), pp. 395-446. Plenum Press, New York.
Hoffman, C.S. and F. Winston. 1987. A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids
for transformation of Escherichia coli. Gene 57: 267-272.
--.
1989. A transcriptionally regulated expression vector for
the fission yeast Schizosaccharomyces pombe. Gene 84:
473-479.
--.
1990. Isolation and characterization of mutants constitutive for expression of the [bpl gene of Schizosaccharomyces pombe. Genetics 124: 807-816.
Ito, H., Y. Fukuda, K. Murata, and A. Kimura. 1983. Transformation of intact yeast cells treated with alkali cations. ].
Bacteriol. 153: 163-168.
Johnston, M. 1987. A model fungal gene regulatory mechanism:
The GAL genes of Saccharomyces cerevisiae. Microbiol.
Rev. 51: 458-476.
Kawamukai, M., K. Ferguson, M. Wigler, and D. Young. 1991.
Genetic and biochemical analysis of the adenylyl cyclase of
Schizosaccharomyces pombe. Cell regul. {in press).
Kikuchi, Y., Y. Kitazawa, H. Shimatake, and M. Yamamoto.
1988. The primary structure of the leul ÷ gene of Schizosaccharomyces pombe. Curr. Genet. 14: 375-379.
Kohli, J. 1987. Genetic nomenclature and gene list of the fission
yeast Schizosaccharomyces pombe. Curt. Genet. 11: 575589.
Lamph, W.W., V.J. Dwarki, R. Ofir, M. Montminy, and I.M.
Verma. 1990. Negative and positive regulation by transcription factor cAMP response element-binding protein is modulated by phosphorylation. Proc. Natl. Acad. Sci. 87: 43204324.
Levitzki, A. 1988. From epinephrine to cyclic AMP. Science
241: 800-806.
Maeda, T., N. Mochizuki, and M. Yamamoto. 1990. Adenylyl
cyclase is dispensible for vegetative cell growth in fission
yeast Schizosaccharomyces pombe. Proc. Natl. Acad. Sci.
87: 7814-7818.
Magasanik, B. 1961. Catabolite repression. Cold Spring Harbor
Syrup. Quant. Biol. 26: 249-256.
Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecudar
cloning: A laboratory manuaI. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.
Matsumoto, K., I. Uno, Y. Oshima, and T. Ishikawa. 1982a.
Isolation and characterization of yeast mutants deficient in
adenylate cyclase and cAMP-dependent protein kinase. Proc.
Natl. Acad. Sci. 79: 2355-2359.
Matsumoto, K., I. Uno, A. Toh-e, T. Ishikawa, and Y. Oshima.
1982b. Cyclic AMP may not be involved in catabolite repression in Saccharomyces cerevisiae: Evidence from mutants
capable of utilizing it as an adenine source. J. Bacteriol.
150: 277-285.
Matsumoto, K., I. Uno, T. Ishikawa, and Y. Oshima. 1983. Cyclic AMP may not be involved in catabolite repression in
Saccharomyces cerevisiae: Evidence from mutants unable
Downloaded from www.genesdev.org on September 10, 2007
Repression of fbpl transcription by cAMP
to synthesize it. J. Bacteriol. 156: 898-900.
Mbonyi, K., M. Beullens, K. Detremerie, L. Geerts, and J.M.
Thevelein. 1988. Requirement of one functional RAS gene
and inability of an oncogenic ras variant to mediate the glucose-induced cyclic AMP signal in the yeast Saccharomyces
cerevisiae. Mol. Cell. Biol. 8: 3051-3057.
McLeod, M. 1989. Regulation of meiosis: From DNA binding
protein to protein kinase. Bioessays 11: 9-14.
Messing, J. 1979. A multipurpose cloning system based on the
single-stranded bacteriophage M13. Recomb. DNA Tech.
Bull. 1: 43-44.
Nadin-Davis, S.A., R.C.A. Yang, S.A. Narang, and A. Nasim.
1986a. The cloning and characterization of a RAS gene from
Schizosaccharomyces pombe. J. Mol. Evol. 23: 41-51.
Nadin-Davis, S.A., A. Nasim, and D. Beach. 1986b. Involvement of ras in sexual differentiation but not in growth control in fission yeast. EMBO J. 5: 2963-2971.
Nadin-Davis, S.A., A. Nasim, Y. Fukui, and M. Yamamoto.
1989. Oncogene homologs. In Molecular biology of the fission yeast (ed. A. Nasim, P. Young, and B.F. Johnson), pp.
97-126. Academic Press, New York.
Nikawa, J., S. Cameron, T. Toda, K.M. Ferguson, and M. Wigler.
1987. Rigorous feedback control of cAMP levels in Saccharomyces cerevisiae. Genes & Dev. 1: 931-937.
Robinson, L., J. Gibbs, M. Marshall, I. Sigal, and K. Tatchell.
1987. CDC25: A component of the RAS-adenylate cyclase
pathway in Saccharomyces cerevisiae. Science 235: 12181221.
Roesler, W.J., G.R. Vandenbark, and R.W. Hanson. 1988. Cyclic
AMP and the induction of eukaryotic gene transcription. J.
Biol. Chem. 263: 9063-9066.
Saloman, Y., C. Londos, and M. Rodbell. 1974. A highly sensitive adenylate cyclase assay. Anal. Biochem. 58: 541-548.
Smith, P.K., R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H.
Gartner, M.D. Provenzano, E.K. Fujimoto, N.M. Goeke, B.J.
Olsen, and D.C. Klenk. 1985. Measurement of protein using
bicinchoninic acid. Anat. Biochem. 150: 76--85.
Sherman, F., G.R. Fink, and C.W. Lawrence. 1978. Cold Spring
Harbor Laboratory manual. Methods in yeast genetics, Cold
Spring Harbor Laboratory, Cold Spring Harbor, New York.
Southern, E.M. 1975. Detection of specific sequences among
DNA fragments separated by gel electrophoresis. J. Mol.
Biol. 98: 503-517.
Tanaka, K., K. Matsumoto, and A. Toh-e. 1989. IRAI: An inhibitory regulator of the RAS/cyclic AMP pathway in Saccharomyces cerevisiae. Mol. Cell. Biol. 9: 757-768.
Tanaka, K., M. Nakafuku, F. Tamanoi, Y. Kaziro, K. Matsumoto, and A. Toh-e. 1990. IRA2, a second gene of Saccharomyces cerevisiae that encodes a protein with a domain
homologous to mammalian ras GTPase-activating protein.
Mol. Cell. Biol. 10: 4303-4313.
Taylor, W.E. and E.T. Young. 1990. cAMP-dependent phosphorylation and inactivation of yeast transcription factor ADR1
does not affect DNA binding. Proc. Natl. Acad. Sci.
87: 4098-4102.
Thevelein, J.M. 1988. Regulation of trehalase activity by phosphorylation-dephosphorylation during developmental transitions in fungi. Exp. Mycol. 12: 1-12.
Toda, T., I. Uno, T. Ishikawa, S. Powers, T. Kataoka, D. Broek,
S. Cameron, J. Broach, K. Matsumoto, and M. Wigler. 1985.
In yeast, RAS proteins are controlling dements of adenylate
cyclase. Cell 40: 27-36.
Toda, T., S. Cameron, P. Sass, M. Zoller, J.D. Scott, B. McMullen, M. Hurwitz, E.G. Krebs, and M. Wigler. 1987a.
Cloning and characterization of BCY1, a locus encoding a
regulatory subunit of the cAMP dependent protein kinase in
Saccharomyces cerevisiae. Mol. Cell. Biol. 7: 1371-1377.
Toda, T., S. Cameron, P. Sass, M. Zoller, and M. Wigler. 1987b.
Three different genes in S. cerevisiae encode the catalytic
subunits of the cAMP-dependent protein kinase. Cell
50: 277-287.
Vassarotti, A. and J.D. Friesen. 1985. Isolation of the fructose1,6-bisphosphatase gene of the yeast Sehizosaccharomyces
pombe. J. Biol. Chem. 260: 6348-6353.
Watanabe, Y., Y. Iino, K. Furuhata, C. Shimoda, and M. Yamamoto. 1988. The S. pombe mei2 gene encoding a crucial
molecule for commitment to meiosis is under the regulation
of cAMP. EMBO J. 7: 761-767.
Wright, A., K. Maundrell, W.D. Heyer, D. Beach, and P. Nurse.
1986. Vectors for the construction of gene banks and the
integration of cloned genes in Schizosaccharomyces pombe
and Saccharomyces cerevisiae. Plasmid 15: 156-158.
Yamawaki-Kataoka, Y., T. Tamaoki, H.-R., Choe, H. Tanaka,
and T. Kataoka. 1989. Adenylate cyclases in yeast: A comparison of the genes from Schizosaccharomyces pombe and
Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. 86: 56935697.
Young, D., M. Riggs, J. Fields, A. Vojtek, D. Brock, and M.
Wiglet. 1990. The adenylyl cyclase gene from Schizosaccharomyces pombe. Proc. Natl. Acad. Sci. 86: 7989-7993.
GENES & DEVELOPMENT
571